Research
Published papers
Planning sustainable urban lighting for biodiversity and society, Nature Cities, 2025, with Léa Tardieu, Sarah Potin, Julie Chaurand, Léa Mariton, Vincent Delbar and Maia David. Pre-print version. Associated R-Shiny tool
Abstract
Urban planners continuously face the challenge of reducing artificial lighting to protect biodiversity while ensuring urban residents’ comfort and safety at night. Striking this balance is crucial for supporting urban residents broadly, yet it remains insufficiently explored in current research. Here we integrate remote sensing and ecological modeling to assess species’ requirements around light-pollution reduction with socioeconomic modeling to evaluate human residents’ acceptance of various street-lighting adjustments, aiming to identify the optimal lighting compromises for Montpellier, France, a midsized European city. We show that, depending on the spatial context, both tradeoffs and synergies can emerge when implementing light-pollution-mitigation measures. By integrating results into an RShiny application, we enabled urban planners to prioritize actions for each streetlight. Our findings underscore the importance of tailoring lighting policies to the specific environmental and social context rather than adopting a universal ‘one-size-fits-all’ approach.
Are citizens willing to accept changes in public lighting for biodiversity conservation?, Ecological Economics, 2022, 200, pp.107527, with Léa Tardieu and Maia David. Pre-print version
Abstract
Light pollution has significantly increased in recent years, in concert with urban sprawl. Light pollution consequences for nocturnal wildlife, human health, and energy consumption are numerous but are poorly tackled in urban policies. The regulation and mitigation of light pollution is possible, but requires an important shift in the lighting paradigm, including in public lighting often managed by local authorities. One of the main sources of reticence of local authorities to regulate light pollution is the potential rejection by citizens of lighting changes. In this article, we investigate citizens’ willingness to accept the transition to more sustainable lighting regimes. We use a discrete choice experiment in a large French metropolis to measure the relative weight of different characteristics of public lighting – light intensity, light extinction, light colour – in respondents’ decisions. We show that respondents are globally open to public lighting shifts, but their preferences in terms of the changes are highly heterogeneous. By incorporating socioeconomic variables of respondents into our econometric models, we characterise the main profiles of preferences regarding lighting changes. This provides practical information to urban and environmental planners allowing them to match the municipalities where the need for light pollution control is a priority with those where measures seem socially acceptable by citizens.
From Poverty to Disaster and Back: a Review of the Literature, Economics of Disasters and Climate Change, 2020, vol. 4, p. 223-247, with Stéphane Hallegatte, Adrien Vogt-Schilb, Julie Rozenberg and Mook Bangalore
Abstract
Poor people are disproportionally affected by natural hazards and disasters. This paper provides a review of the multiple factors that explain why this is the case. It explores the role of exposure (often, but not always, poor people are more likely to be affected by hazards), vulnerability (when they are affected, poor people tend to lose a larger fraction of their wealth), and socio-economic resilience (poor people have a lower ability to cope with and recover from disaster impacts). Finally, the paper highlights the vicious circle between poverty and disaster losses: poverty is a major driver of people’s vulnerability to natural disasters, which in turn increase poverty in a measurable and significant way. The main policy implication is that poverty reduction can be considered as disaster risk management, and disaster risk management can be considered as poverty reduction.
Work in progress
Mapping preferences derived from a choice experiment, with Léa Tardieu, Romain Crastes dit Sourd and Maia David. Working Paper. Under review in Environmental and Resource Economics.
Abstract
In recent decades, discrete choice experiments (DCEs) have proven useful for guiding policy-making decisions, particularly regarding the expression of individual preferences for various policy options. However, when considering spatial planning policies, the usefulness of results such as averaged individual or group preferences may prove limited for decision-makers, since the spatial heterogeneity of preferences is strong in most cases. This paper therefore proposes two methods for mapping, at a small spatial scale, preferences derived from a DCE. Both methods assess the influence of socio-demographic and spatial variables on the preferences to predict willingness to pay at a small spatial scale. Monte Carlo simulations are used to determine the method which best performs in theory. The methods are then compared empirically in the case of light pollution mitigation policies in the Montpellier Metropolitan Area (France). We conclude that, when applicable, the one-step method is the most fit for the purpose, although the two-step method shows practical aspects that can be of interest for applied research.
Identifying public lighting switch-offs from space: A breakpoint detection approach using VIIRS data, with Christopher Kyba and Sébastien Castets.
Abstract
Recently, public lighting switch-off measure gained visibility and scope in some European countries including France. Although such initiatives were already underway in rural areas for various reasons (i.e. biodversity conservation, energy savings), the 2022 energy crisis, consecutive to the war in Ukraine, accelerated their adoption across the country. Yet, since there is no centralised information on the adoption, a rigorous evaluation of the implementation of the measure at the national level, or its variations over time, is impossible. This paper presents a new methodology to detect and date public lighting switch-offs in French municipalities using VIIRS nighttime radiance satellite data from 2012 to 2023. Specifically, we develop an algorithm that detects breaks in radiance time series and further use a random forest algorithm to classify the breaks and distinguish true lighting switch-offs from other radiance changes. Using a database containing the actual information, we obtain a classification accuracy of 88.6\%. The resulting dataset covers all French mainland municipalities with more than 1,500 inhabitants. Our findings reveal that 64.4% of municipalities implemented lighting switch-offs during the study period, including 53.5% after July 2022, when electricity prices were high. This confirms that economic considerations played a significant role in motivating municipalities to adopt these measures. Then, we analyze the characteristics of municipalities adopting these policies, and find that less population, less dense and left-wing municipalities are more likely to switch-off public lighting. The algorithm and dataset developed in this paper offer a robust foundation for evaluating the environmental, social, and economic impacts of lighting reduction policies.
Assessing the impact of public lighting switch-off policies on crime.
Non peer-reviewed articles and opinion papers
Early Warnings: Coverage, Determinants of Reception, and Benefits. What surveys can tell us. World Bank, Policy Research Working Paper 11199 (2025), with Louise Bernard, Bramka Arga Jafino and Paolo Avner.
Éclairer la ville ou protéger la biodiversité : faux dilemme in The Conversation (2025), with Léa Tardieu, léa Mariton, and Maia David
Éclairage public : les Français sont-ils prêts à éteindre la lumière ? in The Conversation (2022), with Léa Tardieu, Maia David, Christine Jez and Jennifer Amsallem
Leveling Up - Impacts of Performance-Based Grants on Municipal Revenue Collection in Mozambique World Bank, Policy Research Working Paper 9789 (2021), with Alvina Erman and Carla Solis Uehara